Dynamic Programming Of Advanced Nanometer Flash Memory

Abstract

An improved method and apparatus for programming advanced nanometer flash memory cells is disclosed.

Description

PRIORITY CLAIM

[0001] This patent application claims priority to and is a divisional of U.S. patent application Ser. No. 13/830,207, filed on Mar. 14, 2013, and titled "Dynamic Programming of Advanced Nanometer Flash Memory," which is incorporated herein by reference in its entirety.

TECHNICAL FIELD

[0002] An improved method and apparatus for programming advanced nanometer flash memory cells is disclosed.

BACKGROUND OF THE INVENTION

[0003] Non-volatile semiconductor memory cells using a floating gate to store charges thereon and memory arrays of such non-volatile memory cells formed in a semiconductor substrate are well known in the art. Typically, such floating gate memory cells have been of the split gate type, or stacked gate type.

[0004] The prior art includes a common technique for programming a flash memory cell. In the prior art technique, a high voltage is applied on the drain of the memory cell, a bias voltage on the control gate of the memory cell, and a bias current on the source of the memory cell. The programming essentially places electrons on the floating gate of the memory cell. This is described in U.S. Pat. No. 7,990,773, "Sub Volt Flash Memory System," by Hieu Van Tran, et al, which is incorporated herein by reference.

[0005] An example of prior art circuitry for providing the bias current applied to the source of each memory cell is shown in FIG. 1. Flash memory system 10 comprises flash memory array 30, horizontal decoder 20, and vertical decoder 50. Horizontal decoder 20 selects the row (often called the word line) within flash memory array 30 to be read or written, and vertical decoder 50 selects the column (often called the bit line) within flash memory array 30 to be read or written. Vertical decoder 50 comprises a series of multiplexers coupled to flash memory array 30. Flash memory array 30 comprises N blocks of cells, and each block is coupled to one multiplexer within vertical decoder 50. Current source 40 is coupled to N current mirrors, current minor 60₁ through current minor 60_N.

[0006] One drawback of the prior art method of programming flash memory cells is that current minors, such as current mirrors 60₁ through 60_N, often are mismatched due to natural variations and manufacturing variances, and in large chips, the ground potential also might vary. As a result, during operation the current minors actually may draw less or more current than they are supposed to draw.

[0007] What is needed is an improved method and apparatus for programming flash memory cells, particularly advanced nanometer flash memory cells, that reduces or eliminates the variability among the bias current sources used during the programming process.

SUMMARY OF THE INVENTION

[0008] The aforementioned problems and needs are addressed through two different embodiments.

[0009] In one embodiment, two or more current mirrors are averaged together to draw current from each block in the memory array during the programming process, instead of using just one current mirror. The end result is less variability in the current drawn in programming each cell.

[0010] In another embodiment, each current mirror is initialized prior to operation, and a capacitor is used to keep the current mirror at or near the initialized state.

BRIEF DESCRIPTION OF THE DRAWINGS

[0011] FIG. 1 depicts a prior art circuit for programming a memory circuit.

[0012] FIG. 2 depicts an embodiment for applying an average of a plurality of current mirrors to a memory array.

[0013] FIG. 3 depicts a timing diagram for the activation of switches depicted in FIG. 2.

[0014] FIG. 4 depicts an embodiment of a programming circuit, where each current minor is first initialized before being used to program memory cells.

[0015] FIG. 5 depicts another embodiment of a programming circuit, where each current minor is first initialized before being used to program memory cells.

[0016] FIG. 6 depicts another embodiment of a programming circuit, where a first plurality of current mirrors and a second plurality of current mirrors are alternately initialized and used to program memory cells.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] In the prior art system of FIG. 1, the current drawn by current minors 60₁ . . . 60_N can vary substantially in the real world. For example, if each current minor ideally should draw 1.00 μA, the inventors have found that in the real world, some typical measured values might be as shown in Table 1:

TABLE-US-00001 TABLE 1 Current Mirror Current IB0 1.32 μA IB1 0.64 μA IB2 1.62 μA IB3 0.71 μA

[0018] Table 1 depicts four exemplary current minors, labeled IB0, IB1, IB2, and IB3. As can be seen, the variance in current is significant. One way to compensate for the variance is to use averages of the current minors. For example, the average of IB0 and IB is 0.98 μA, the average of IB2 and IB3 is 1.165 μA, and the average of IB0, IB1, IB2, and IB3 is 1.0725 μA. Based on this simple example, the deviation in current from the ideal of 1.00 μA is reduced from 62% for individual current mirrors to 16.5% for averaging two current mirrors and 7.3% for averaging four current mirrors.

[0019] With this observation in mind, an embodiment is now described with reference to FIG. 2. Programming circuit 100 comprises current minors 110, 111, 112, and 113 and bit lines 160, 170, 180, and 190. Current mirror 110 can be coupled to bit lines 160, 170, 180, and 190 through switches 120, 130, 140, and 150, respectively. Similarly, current minor 111 can be coupled to bit lines 160, 170, 180, and 190 through switches 121, 131, 141, and 151, respectively; current mirror 112 can be coupled to bit lines 160, 170, 180, and 190 through switches 122, 132, 142, and 152, respectively; and current mirror 113 can be coupled to bit lines 160, 170, 180, and 190 through switches 123, 133, 143, and 153, respectively. In this design, controller 200 creates control signals 201, 202, 203, and 204 to control the switches. For example, switches 120, 121, 122, and 123 are controlled by control signal 201, 202, 203, and 204 respectively; switches 130, 131, 132, and 133 are controlled by control signal 204, 201, 202, and 203 respectively; switches 140, 141, 142, and 143 are controlled by control signal 203, 204, 201, and 202 respectively; and switches 150, 151, 152, and 153 are controlled by control signal 202, 203, 204, and 201 respectively.

[0020] In FIG. 3, an example of control signals 201, 202, 203, and 204 generated by controller 200 are shown. Here, a high voltage represents that the switches controlled by the control signal are turned "on," and a low voltage represents that the switches controlled by the control signal are turned "off." FIG. 3 depicts two periods of programming of flash memory cells, such as advanced nanometer flash memory cells. During the first period, control signal 201 is high for a period t_o, then control signal 202 is high for a period t_o, then control signal 203 is high for a period t_o, and then control signal 204 is high for a period t_o. As shown in FIG. 3, there is some intentional overlap between the high voltage occurrences of the control signals.

[0021] The end result is that bit lines 160, 170, 180, and 190 each are coupled to current minors 110, 111, 112, and 113 in equal amounts during a single programming period. This effectively averages the current drawn from each bit line across current mirrors 110, 111, 112, and 113. As discussed above, this temporal averaging effect will reduce the variation in programming current based on mismatches and other variations in the current mirrors.

[0022] One of ordinary skill in the art will appreciate that the example of FIGS. 2 and 3 can be applied to an entire array of flash memory cells and corresponding current minors. One of ordinary skill in the art also will appreciate that the same principles can be used to average any number of current minors instead of four current minors. For example, two current mirrors, eight current minors, or any other number of current mirrors could be averaged over one programming period using the same principles described with reference to FIGS. 2 and 3.

[0023] Another embodiment is now described with reference to FIG. 4. Dynamic programming circuit 300 comprises current source 310 and memory cell replica 320. Memory cell replica 320 is of the same design as the memory cells in the flash memory array in question (not shown). The gate of exemplary current mirror 340 is coupled to capacitor 350. In another embodiment there is no capacitor 350, instead the gate capacitance of the current minor 350 serves as an acting capacitor. Capacitor 350 is coupled to current source 310 through switch 330, and the drain of current minor 340 is coupled to memory cell replica 320. The memory cell replica 320 has a gate bias 360 emulating the gate programming voltage of the memory cell. As shown the current mirror 340 is enhancement NMOS transistor. Alternatively, the current minor 340 can be PMOS transistor.

[0024] During an initial programming mode, switch 330a is turned "on" and capacitor 350a is connected to current source 310, and the drain of current mirror 340a is coupled to memory cell replica 320. Capacitor 350a will be charged until the current through current mirror 340a matches the current of current source 310. Current source 310 is chosen to generate the ideal level of current that is desired for current mirror 340 and other current mirrors, such as 1.0 μA. The memory cell replica 320 with its gate voltage 360 forces a replica drain programming voltage on the drain of the current mirror 340a. This voltage will significantly be the same as in actual memory cell programming operation.

[0025] After the initial programming mode is complete, switch 330a is turned "off," and thereafter capacitor 35a0 is no longer connected to current source 310 and hence now holding a bias voltage on its terminal (same as gate of the current mirror 340), and the drain of current minor 340 is no longer connected to memory cell replica 320. Now an operational programming mode is applied in which current minor 340 can be used to program one or more memory cells in the memory array. As can be appreciated, all current minors in the system can be programmed in the same manner discussed above for current minor 340. Alternatively the initial programming mode (initialization programming period) can be applied sequentially for all the current mirrors 340a to 340n. Then in the operational programming mode all current mirrors 340a-n are applied to selected bitlines of selected memory cells for the programming operation at the same time.

[0026] Thus, the embodiment of FIG. 4 creates a way to initialize the current mirrors of a system to cause them to operate in a near-ideal state with exact programming current with ideal drain programming voltage.

[0027] A variation of the embodiment of FIG. 4 is now shown in FIG. 5 as programming circuit 400. The same devices previously shown in FIG. 4 are shown in FIG. 5 as well and will not be described again for efficiency's sake. However, in FIG. 5, current source 310 couples directly to switch 330a, and switch 330a couples to memory cell replica 320a, and memory cell replica 320a then is coupled to the drain of current mirror 340a. The circuit will otherwise operate the same as in FIG. 4 and the current mirrors can be programmed such that they initially operate in a near-ideal state during operational mode.

[0028] The programming of current minors in the systems of FIGS. 4 and 5 are controlled by controller 390. Controller 390 will control switch 330a-n and the other switches and also will ensure that current minor 340a-n and the other current mirrors are not used for operation of the memory array during the programming mode.

[0029] During the operational mode of the embodiments for FIGS. 4 and 5, the charge of capacitor 350 initially will keep current mirror 340 operating in a state where it draws the same level of current as current source 310. This effect will diminish over time as capacitor 350 loses its charge through normal leakage. Once the charge of capacitor is completely lost, then current minor 340 will operate as it would have operated had the programming mode never occurred.

[0030] In response to this challenge, a further embodiment is shown in FIG. 6. FIG. 6 depicts another dynamic programming circuit 500. Dynamic programming circuit comprises a first plurality 510 of programmable current mirrors and a second plurality 520 of programmable current mirrors. The first plurality 510 of current mirrors and second plurality 520 of current minors can be based upon one or both of the embodiments of FIGS. 4 and 5. When the first plurality 510 is being programmed (in programming mode), the second plurality 520 is used for actual operation (operational mode) of the flash memory array (not shown). After time t₁ elapses, the first plurality 510 is then used for actual operation of the flash memory array (operational mode), and the second plurality 520 is programmed (programming mode). After another elapse of time t₁, the first plurality enters programming mode and the second plurality 510 enters operation mode, and so on. Time t₁ is selected to be a time interval such that capacitive charge has not yet diminished to such a point that the operation of the associated current mirror has reached an undesirable level. For example, time t₁ might be chosen to be 1 ms in certain systems.

[0031] References to the present invention herein are not intended to limit the scope of any claim or claim term, but instead merely make reference to one or more features that may be covered by one or more of the claims. Materials, processes and numerical examples described above are exemplary only, and should not be deemed to limit the claims. It should be noted that, as used herein, the terms "over" and "on" both inclusively include "directly on" (no intermediate materials, elements or space disposed there between) and "indirectly on" (intermediate materials, elements or space disposed there between). Likewise, the term "adjacent" includes "directly adjacent" (no intermediate materials, elements or space disposed there between) and "indirectly adjacent" (intermediate materials, elements or space disposed there between). For example, forming an element "over a substrate" can include forming the element directly on the substrate with no intermediate materials/elements there between, as well as forming the element indirectly on the substrate with one or more intermediate materials/elements there between.

Claims

1. A programming circuit for use in a memory device, comprising: a current mirror comprising a transistor; a capacitor coupled to the gate of the transistor; and a switch coupled to a current source and the capacitor, wherein the current source is coupled to the capacitor through the switch in a first mode and the current source is uncoupled from the capacitor and is coupled to a memory cell during a second mode.

2. The circuit of claim 1, wherein the memory cell is a flash memory cell.

3. The circuit of claim 1, wherein the memory cell is a split gate flash memory cell.

4. The circuit of claim 1, wherein the transistor is an NMOS transistor.

5. A programming circuit for use in a memory device, comprising: a current mirror comprising a transistor; and a switch coupled to a current source and the gate of the transistor, wherein the current source is coupled to the gate of the capacitor through the switch in a first mode and the current source is uncoupled from the gate of the capacitor and is coupled to a memory cell during a second mode.

6. The circuit of claim 5, wherein the memory cell is a split gate flash memory cell.

7. A method for programming a memory device, comprising: coupling a current source to a capacitor through a switch during a first mode; decoupling the current source from the capacitor during a second mode; providing charge from the capacitor to the gate of a transistor during the second mode; drawing current from a memory cell coupled to the drain of the transistor during the second mode.

8. The method of claim 7, wherein the memory cell is a flash memory cell.

9. The method of claim 7, wherein the transistor is an NMOS transistor.

10. The circuit of claim 8, wherein the memory cell is a split gate flash memory cell.

11. A programming circuit for use in a memory device, comprising: a plurality of current mirrors; a bit line coupled to a plurality of memory cells and to a set of switches, each of the switches in the set of switches coupled to one of the plurality of current minors; a controller configured to control the set of switches during a programming operation of one of the plurality of memory cells, wherein each of the switches in the set of switches is turned on and turned off during at least part of the programming operation.

12. The circuit of claim 11, wherein the set of switches comprises two switches.

13. The circuit of claim 11, wherein the set of switches comprises four switches.

14. The circuit of claim 11, wherein the set of switches comprises eight switches.

15. The circuit of claim 11, wherein during the programming operation the switches in the set of switches are turned on and off sequentially.

16. The circuit of claim 13, wherein two or more switches of the set of switches are on at the same time during at least part of the programming operation.

17. The circuit of claim 11, wherein the controller causes an average of currents through each of the plurality of current mirrors to be provided to the bit line during the programming operation.

18. The circuit of claim 11, wherein the memory cells are flash memory cells.

19. A method of programming a memory device comprising a plurality of current minors, a bit line coupled to a plurality of memory cells and to a set of switches, each of the switches in the set of switches coupled to one of the plurality of current minors, and a controller configured to control the set of switches during a programming operation of one of the plurality of memory cells, the method comprising: turning on and off, by the controller during the programming operation, each switch in the set of switches in sequence.

20. The method of claim 19, wherein the set of switches comprises two switches.

21. The method of claim 19, wherein the set of switches comprises four switches.

22. The method of claim 19, wherein the set of switches comprises eight switches.

23. The method of claim 19, wherein two or more switches of the set of switches are on at the same time during at least part of the programming operation.

24. The method of claim 23, wherein during the programming operation, the controller causes an average of currents through each of the plurality of current minors to be provided to the bit line.

25. The method of claim 19, wherein the memory cells are flash memory cells.

Images